The blissful Zen of a good side project

I love this.

My most satisfying side projects are often not necessarily my "best" work, in terms of code cleanliness, best practices, efficiency, etc. They're ones where I had a particular creative itch I wanted to scratch. Is this kind of solution possible? What would a certain unusual approach to a problem look like? How can I use this algorithm or library in this situation where it doesn't quite fit, as an experiment?

Projects with extremely loose parameters and no particular "skill acquisition" goals are great ways to grow in ways you didn't anticipate. Which is one way to think about artistic creation, I think: non-goal oriented growth.

I agree with that last bit - sometimes you gotta drop the time associated with polish of a finished product and play around.

Always stuck with me that pretty much every famous piece of art has a long backlog of practice to get to that point.